Currently,, in some specific working conditions, the demand for control accuracy of the welding inverter soft switch is very high, such as accuracy and efficiency and other characters. But it is difficult to get desirable effect with traditional PID controller. Therfore the particle swarm neural network algorithm is adopted which has higher control accuracy and the PID controller with self-adjustment function is designed. This algorism intelligently adjusts the three parameters of PID through the online learning of system state, controls the on-off time of the inverter ' s power switches and finally helps to realize the control of constant currency output of the whole system. Simulation results show that the algorithm improves the control accuracy of welding inverter soft switch, the system' s robustness and self-adaptiveness was also improved.%目前,在一定的工业生产中,要求逆变软开关弧焊电源具有精确、有效的控制.但传统中使用的的PID控制器的控制效果很难达到令人满意的程度.因此,采用粒子群神经网络算法,设计了具有自适应调整功能的PID控制器.通过对系统运行状态的在线学习,智能化地修正PID的三个输出参数,进而控制系统中逆变器的功率开关器件的导通时间,最终实现整个系统的恒电流输出控制.仿真结果表明:将粒子群神经网络算法运用到逆变软开关弧焊电源中,可使系统控制效果得到很大改善.同时使整个系统具有更好的自适应性和鲁棒性.
展开▼
